Swarthmore College Swarthmore College

Works Works

Philosophy Faculty Works Philosophy

11-1-1972

Review Of "A Theory Of Justice" By J. Rawls Review Of "A Theory Of Justice" By J. Rawls

Hans Oberdiek Swarthmore College, hoberdi1@swarthmore.edu

Follow this and additional works at: https://works.swarthmore.edu/fac-philosophy

Part of the Philosophy Commons

Let us know how access to these works benefits you

Recommended Citation Recommended Citation Hans Oberdiek. (1972). "Review Of "A Theory Of Justice" By J. Rawls". New York University Law Review. Volume 47, Issue 5. 1012-1022. https://works.swarthmore.edu/fac-philosophy/118

This work is brought to you for free by Swarthmore College Libraries' Works. It has been accepted for inclusion in Philosophy Faculty Works by an authorized administrator of Works. For more information, please contact myworks@swarthmore.edu.

BOOK REVIEWS

A THEORY or JusTicE. By John Rawls. Cambridge, Massachusetts:Harvard University Press. 1971. Pp. xv, 607. $15.00 cloth, $3.95 paper.*

IINTRODUCTION

The contribution of twentieth-century moral philosophy to thesolution of pressing social problems has been minimal. G.E. Moore's"naturalistic fallacy" and the application of A.J. Ayer's verifiabilitycriterion of meaning convinced a generation of philosophers that ethics,understood as the formulation and defense of substantive moral prin.ciples, is impossible. Philosophers' only legitimate task, so it wasthought, lay in elucidating the language of morality and the vocabularyof politics. Thus, in a well-known book intended to bring the reader togrips with the fundamental problems of ethics, the author declared,"Ethics, as I conceive it, is the logical study of the language of morals."'

Ethics, paradoxically, seemed to be irrelevant to moral concerns. Oneneed not deny that linguistic analysis has clarified and deepened ourunderstanding of moral concepts and judgments to suggest that phi-losophers must do more. In A Theory of Justice, John Rawls demon-strates that one can address living moral, political and social issuessensitively and forcefully without sacrificing philosophic standards ofclarity and rigor.

A Theory of Justice includes much more that its title suggests. Inaddition to elaborating and defending principles of social justice, moralsentiments, principles of moral psychology and the unity of self, thenature of goodness and happiness are discussed. While these and othersubjects are not simply addenda to Rawls's theory, I shall concentratesolely on the theory of social justice developed in the first two-thirds of

* Citations to A Theory of Justice are to the paperback edition.1 R. Hare, The Language of Morals, at v (1952). Hare is not alone in his

preoccupation with meta-ethical questions-with questions about moral discourse,especially questions concerning meaning, the nature of moral judgments andmethods of justifying or supporting moral judgments. Throughout the nineteenforties and fifties, substantive moral and political philosophy was virtually dead.Studies sharing Hare's conception of ethics are: K. Baler, The Moral Point ofView (1958); P. Nowell-Smaith, Ethics (1954); C. Stevenson, Ethics and Language(1943); and S. Toulmin, Reason in Ethics (1990). For an instructive essay, seeIV. Frankena, Moral Philosophy at Mid-Century, 60 Philosophical Rev. 44(1951). Two questions have troubled writers on meta-ethics in recent years: (1)Does acceptance of a particular meta-ethical theory logically commit one to theacceptance of any particular normative theory, and the converse; and (2) Is thedistinction between normative and meta-ethical judgments sound? Tie firstquestion concerns independence; the second concerns neutrality. One of tice firstbooks to break away from an exclusive concern with mcta-etbics is M. Singer,Generalization in Ethics (1961). Curiously, Rawls, who is generous In citing hispredecessors, never mentions Singer's book.

Imaged with the Permission of N.Y.U. Law Review

BOOK REVIEWS

the book. Particular attention will be paid to Rawls's general approachto theory-construction in ethics, his contractualism and the novel ana-lytic construct from which he derives his twin principles of justice. Onlyafter examining Rawls's methodology will I consider the principles ofjustice themselves. So that we will not lose our bearings, however, Ibegin with a thumbnail sketch of Rawls's general argument.

All of us have intuitive beliefs regarding justice, though few of ushave anything like a theory of justice. Because our intuitions conflict orotherwise fail us, we seek a theory which will systematize and deepenour conception of justice. In developing any theory, we cannot beginfrom scratch; we must begin with our considered judgments, judgmentsthat any plausible principle must confirm. However, as our theorydevelops we may modify or even abandon some of our initial beliefs,especially those which are, upon reflection, confused, inconsistent orthe consequence of prejudice. Yet, however far our theory takes us fromthese initial beliefs, we cannot begin without them. Our aim is to reacha state of "reflective equilibrium," a state in which our principles andjudgments coincide and in which we know the principles to which ourjudgments conform and the premises of their derivation (cf. p. 20).

To guide our deliberations we may "imagine that those who engagein social cooperation choose together, in one joint act, the principleswhich are to assign basic rights and duties and to determine the divisionof social benefits" (p. 11). The "most philosophically favored inter-pretation" of this initial choice situation, called the original position,will show that two principles of justice would be agreed upon, the firsttaking precedence over the second:

First: each person is to have an equal right to the most extensivebasic liberty compatible with a similar liberty for others.

Second: social and economic inequalities are to be arranged so thatthey are both (a) reasonably expected to be to everyone's advantage,and (b) attached to positions and offices open to all (p. 60).

These principles apply primarily to the basic structure of society, i.e.,to the political constitution and to the principal economic and socialinstitutions. They may or may not have wider application. Furtherboth principles assume full compliance; social problems arising out ofpartial compliance create special difficulties. However, only an idealtheory can provide a systematic basis for the solution of contemporarysocial problems; unless we understand the nature and aims of a per-fectly just society, we will lack a satisfactory standpoint from which toevaluate existing institutions and their practices.

II

METHODOLOGY

A. The Pragmatic Approach

Rawls adopts a currently fashionable account of theory-constructionin the sciences. Rather than theorize on a bedrock of indubitable truths,

Imaged with the Permission of N.Y.U. Law Review

NEW YORK UNIVERSITY LAW REVIEW [Vol. 47:1012

as Cartesians would have us do, we are urged to begin with what we donot in fact doubt, e.g., that water boils at 100' C. at sea level. Hy-potheses are then formed to ground such beliefs and to shed light on avariety of theoretical and practical problems we wish to solve (such aswhy altitude affects the boiling point of water). At an early stage oftheorizing, hypotheses usually give way to firmly held beliefs if conflictsbetween them occur; but as hypotheses become imbedded in our con.ceptual scheme because of their explanatory power and connections withother hypotheses, we often sacrifice cherished beliefs in order to save ourtheories. In short, theorizing is a dialectical, self-correcting process ofmutual adjustment among observations, hypotheses and theories.

Physical theories are now deeply imbedded in our thinking, thoughthis was not always so. Psychological and sociological theories, on theother hand, have significantly less standing, so we are reluctant tomodify our intuitive beliefs about human nature and society to satisfysomeone's pet hypothesis. Ethical theories have even less claim to ourallegiance. Most of us, for instance, believe that racism and slaveryare unjust; if any theory implies otherwise, so much the worse for itlWhat we seek is a theory that will explain the injustice of racism andslavery and at the same time provide guidance where our moral beliefsare vague, tentative or inconsistent. While it would be rash to supposethat any ethical theory could possess the power of physical or biologicaltheories, Rawls believes that his own conception of justice as fairness isa distinct improvement over previous conceptions, including utilitarianconceptions.

One virtue of Rawls's pragmatic approach is that it narrows thegap between scientific and ethical theories. Briefly, a scientific theorymust answer adequately serious objections to it, have good reasons inits favor and be superior to its rivals in terms of simplicity, fruitfulness,thoroughness and compatibility with other accepted theories., Thesesame criteria apply to ethical theories, albeit in a somewhat differentfashion. Unless an ethical theory can answer serious objections, forexample, it must be rejected. No doubt the nature of "serious objec-tions" and "adequate answers" changes as we pass from science toethics, but this should not blind us to the underlying unity of thecritical processes involved or to the fact that the nature of "seriousobjections" and "adequate answers" also varies within the sciencesthemselves. Another virtue of Rawls's pragmatic approach is that defi-

2 There is a cursory discussion of the criteria of acceptability in science andtheir applicability to ethics in R. Garner & B. Rosen, Moral Philosophy, ch. 1(1967). For a somewhat lengthier discussion of criteria of confirmation and ac-ceptability, see R. Harr6, An Introduction to the Logic of the Sciences (1960),or C. Hempel, Philosophy of Natural Science (1966). Rawls himself traces hisapproach to the work of IV. Quine, Word and Object (1960), and of M. Whte,Toward Reunion in Philosophy (1956). Quine and White, colleagues of Rawb atHarvard, trace their roots back to the pragmatism of Charles Sanders Peirceo andJohn Dewey.

Imaged with the Permission of N.Y.U. Law Review

BOOK REVIEWS

nitions are kept in their places Key concepts are introduced and ex-plained only as needed and only within the context of the entire theory;they have little or no life away from their theoretical home. Definitionsoccur, if at all, only after one understands the role of the defined con-cept within the theory. Thus, theories and definitions stand or falltogether, an insight not always acknowledged by linguistic analysts.

One may feel that the pragmatic approach is bound to generatecircular theories, since one uses intuitive convictions to formulate prin-ciples that are then confirmed by showing that they account for thesesame convictions. Were this all that was accomplished, one's theorywould indeed be circular; fortunately, Rawls strives for much more.He argues, for example, that adoption of his two principles reconcilesfreedom and reason, helps explain the place of moral sentiments (such asshame, guilt, resentment and indignation) in our lives, fixes the properrelationship between majority rule and minority rights and explainsthe grounds for civil disobedience and political obligation. Principles ofjustice established through Rawls's general pragmatic approach there-fore take us beyond our initial beliefs both by providing guidance inmatters where we lack firm convictions and by showing how apparentlyunrelated beliefs form a coherent system.4

B. The Social Contract and the Original Position

Social contract theories are in general disrepute for reasons which,if not always sound, are at least well-known. Rawls's decision to casthis theory in contractarian language may therefore come as a surprise.However, he sidesteps many familiar objections by denying that thecontracting parties he has in mind are negotiating to enter a particularsociety or to form a particular government. Rather,

3 See G. Field, The Place of Definition in Ethics, in 3 University of BristolStudies, Studies in Philosophy (1935), reprinted in Readings in Ethical Theory92-102 (1st ed. W. Sellars & J. Hospers eds. 1952).

4 The pragmatic approach espoused by Rawls is not without pitfalls andlimitations. William James, for instance, thought that a coherent, fruitful andsuperior theory was true. It follows from James's analysis that the Ptolemaictheory of planetary motion was once true but is now false. Rawls avoids thismistake, though he does so in part because he ignores questions of truth altogether.This may or may not be a virtue of the contract approach. Even granting theusefulness of the pragmatic approach in science, one wonders whether moraljudgments, beliefs and convictions can play the same role that reports ofempirical observations play. This question arises not simply because agreement onmoral matters is difficult to obtain or because of the complezity of human conduct,but because one needs to determine and justify the status of moral judgments.Skeptics will argue that Rawls offers no more than a theory of a-rational prefer-ences. To put the question in Kantiant terms, Rawls must show that moraljudgments are possible before claiming that he has established an objective theoryof social justice. Following Kant further, we may ask whether ethical theoriesmust in fact justify initial moral judgments at all-perhaps we should begininstead with certain moral concepts (such as duty) that even conflicting initialbeliefs embody.

Imaged with the Permission of N.Y.U. Law Review

November 1972]

NEW YORK UNIVERSITY LAW REVIEW [Vol. 47:1012

the guiding idea is that the principles of justice for the basic structure ofsociety are the object of the original agreement. They are the principlesthat free and rational persons concerned to further their own interestswould accept in an initial position of equality as defining the funda-mental terms of their association. These principles are to regulate tillfurther agreements; they specify the kinds of social cooperation thatcan be entered into and the forms of government that can be established.This way of regarding the principles of justice I shall call justice asfairness (p. 11).

Much of A Theory of Justice is devoted to explaining and defending thecontent of this passage.

Classical contract theorists had employed the notion of a state ofnature to show what life would be like were men to live outside ofsociety and to argue from that hypothetical state of affairs that certainfeatures must characterize a just and well-governed society. BecauseHobbes, Locke and Rousseau offered such radically divergent descrip-tions of life in a state of nature, it is not surprising that they defendedequally divergent political principles and institutions. Although theconcept of an original position resembles a state of nature in somerespects, it differs in that it does not even pretend to describe a possiblehistorical situation-the conditions defining the original position couldnot possibly occur. The original position is in fact a hypothetical con-struct from which one can derive principles of justice that are acceptablefrom a moral point of view. Because it plays a crucial role in the deriva.tion of Raws's two principles, its nature and function must be fullyexplored.

The original position is the "most philosophically favored inter-pretation of [the] initial choice situation for the purposes of a theoryof justice" (p. 18). To satisfy ourselves that we have hit upon the mostphilosophically favored interpretation we must limit ourselves to sharedbut weak conditions and match the principles that would be chosen(given those conditions) against our considered moral convictions. Sincewe wish to derive moral principles from the original position, it isessential that the description of the original position be free of moraland value assumptions; otherwise the principles will only appeal tothose already convinced. I shall follow Rawls in considering the originalposition under the headings of the circumstances of justice, the formalconstraints of the concept of right, the veil of ignorance and the ration-ality of the contracting parties.

By the circumstances of justice Rawls means "the normal condi-tions under which human cooperation is both possible and necessary"(p. 126); in the main, his account follows that of Hume. Society, Rawlsmaintains, is a cooperative venture for mutual advantage marked byboth an identity and a divergence of interests; the identity makes sociallife possible and the divergence makes institutions necessary. Amongthe objective circumstances of justice are that individuals live in close

5 See D. Hume, A Treatise of Human Nature, Book III, pt. II, § II (1740),and An Enquiry Concerning the Principles of Morals, § III, pt. I (1751).

Imaged with the Permission of N.Y.U. Law Review

BOOK REVIEWS

proximity to one another, are vulnerable to attack and to having theirinterests blocked by others, are *roughly similar in physical and mentalpowers and live under conditions of moderate scarcity. The subjectivecircumstances of justice include the fact that individuals, though capableof cooperation and a sense of justice, have collectively a variety ofincompatible life-plans which demand conflicting shares of natural andsocial resources for their attainment. Further, men espouse incompatiblephilosophical, religious and political views. All of us, moreover, sufferthe frailties of human nature: shortcomings of knowledge, thought,judgment and an assortment of moral faults.

Thus far, the description depicts life as it is. In rounding out hisaccount, however, Rawls stipulates that the contracting parties arewholly self-interested; he does so in order "to insure that the principlesof justice do not depend on strong assumptions" (p. 129). The fact thatpeople form bonds of affection and take an altruistic interest in theinterests of others can only strengthen adherence to principles adoptedon the assumption of mutual self-interest.

Those in the original position think and act under certain con-straints. Constraints of right hold for the choice of any moral principles,including principles of social justice. Rawls regards them as formalconstraints, though he sees that they are not without moral force, sincethey rule out egoism. He emphasizes that "the conditions are not justi-fied by definition or the analysis of concepts, but only by the reason-ableness of the theory of which they are a part" (p. 131).G Just as theo-retical entities in physics derive their meaning and justification fromthe theory of which they are a part, so too with Rawls's five formalconstraints on the concept of right.

First, all principles must be general-avoid proper names andarbitrary definite descriptions. Second, all principles must be universalin application-they must hold for everyone, thereby ruling out self-contradictory and self-defeating principles. Third, all principles must bepublicly chosen and acknowledged, thus disqualifying private, secret orotherwise esoteric moralities. Fourth, a conception of right must imposean ordering on conflicting claims, a task which Rawls admits is ex-tremely difficult. And fifth, principles agreed to must be regarded as finaland conclusive. Were we able to formulate a complete theory of moral-ity, the totality of principles would not only determine what counts asa relevant consideration, but would assign weights as well. Conclusionsfrom such principles would override self-interest, custom and law. Aconception of right, then, "is a set of principles, general in form anduniversal in application, that is to be publicly recognized as a final

6 In the main, Rawls follows the analysis of K. Baier, The Moral Point ofView, ch. 8 (1958). There are differences, however. The most important is thatBaier attempts to define the moral point of view from the very conception ofmorality; Baier's criteria define what it means to have a moral point of view. It isthis attempt to avoid engaging in substantive, normative inquiry which Rawlsrejects.

Imaged with the Permission of N.Y.U. Law Review

November 19721

NEW YORK UNIVERSITY LAW REVIEW [Vol. 47,1012

court of appeal for ordering the conflicting claims of moral persons"(p. 135).

Rawls introduces the notion of pure procedural justice to insurethat any principles arrived at in the original position will be just. Oneadopts this procedure, he argues, when there is no independent criterionfor the just outcome: "instead there is a correct or fair procedure suchthat the outcome is likewise correct or fair, whatever it is, provided thatthe procedure has been properly followed" (p. 86). A distinctive featureof pure procedural justice is that the procedure must actually be used.In games of chance, any distribution of prizes is just, provided that fairprocedures are actually used. It would be no good to hand out the prizesrandomly before commencing play and then defend the distribution onthe ground that actually playing the game according to purely proce-dural rules might have had exactly the same outcome. Although true,the defense is irrelevant.

Purely procedural justice will yield a just outcome, however, only ifwe assume that the contracting parties are shrouded in a veil ofignorance, a veil nullifying "the effects of specific contingencies whichput men at odds and tempt them to exploit social and natural circum-stances to their own advantage" (p. 136). Any information that wouldenable one to negotiate for principles tailormade to one's own personalor historical circumstances is excluded. Thus, no one knows anythingof one's natural abilities, place in society, aspirations, special psycho-logical propensities, religious views or even the civilization to which onebelongs (cf. p. 137). However, contractors in the original position areonly selectively ignorant. Rawls assumes that they know a great manythings most of us do not know: general facts about human society (e.g.,the circumstances of justice), the basis of social organization, the lawsof psychology, the nature of politics, principles of economics andscientific theories generally (cf. pp. 137-38).

Particular information is restricted so that the parties have no basisfor bargaining in the usual sense-for urging principles especially favor-able to oneself. Since no one knows any distinguishing facts about one-self, there is no way to hold out for self-serving principles. Generalinformation is permissible, on the other hand, "since conceptions ofjustice must be adjusted to the characteristics of the systems of socialcooperation which they are to regulate" (p. 138).

Individuals in the original position also know that they desirecertain primary goods, i.e., those goods normally necessary to or consti-tutive of anyone's life plan and therefore objects of everyone's rationaldesires. Natural primary goods, such as intelligence and health, concernus less than social primary goods, since only the latter are subject todistribution by social and political institutions. Among the most im-portant social primary goods are rights, liberties, powers, opportunities,income, wealth and, above all, the bases of self-respect (cf. p. 62).

Rawls stipulates that individuals in the initial situation are rational,though rationality is construed narrowly as "taking the most effective

Imaged with the Permission of N.Y.U. Law Review

BOOK REVIEWS

means to given ends" (p. 14). This account of rationality, standard ineconomic and social theory, entails that rational individuals have acoherent set of preferences, rank options according to purposes andadopt plans satisfying more, rather than fewer, desires (cf. p. 143). Byeliminating moral preferences from his account of rationality, Rawlsbelieves that principles derived from his amoral construct are more likelyto be accepted. To be sure, Rawls builds into the original position whathe calls "a thin theory of the good" in order "to explain the rationalpreference for primary goods and to explicate the notion of rationalityunderlying the choice of principles in the original position" (p. 397).A thin theory of good "is necessary to support the requisite premisesfrom which the principles of justice are derived" (id.). Because the thintheory is confined to bare essentials that no one can deny, principlesderived from these essentials are not in danger. Rawls does depart fromthe standard economic account of rationality in one respect: he assumesthat individuals in the original position, or Rational Individuals as Ishall refer to them, are not envious-Individuals who have gained afavorable position at no cost to other Individuals will not be viewedwith hostility. Were envy permitted, the original position would yieldunacceptably egalitarian principles.

Explicit in Rawls's account of rationality is the belief that Indi-viduals have ends, though not that mankind has any specific end. Pur-suit of any end must be regulated by those principles accepted in theoriginal position on the assumption that Individuals are wholly self-interested and have a variety of final ends. This is part of what Rawlsmeans when he says, "in justice as fairness the concept of right is priorto that of the good" (p. 396). To use standard jargon, justice as fair-ness is a deontological, not a teleological, theory.7 Rawls concludes thatthose principles which Rational Individuals would accept in the originalposition are the correct principles of social justice that we, as flesh andblood individuals, should apply to the basic structure of our (or any)society.

Professor Rawis's characterization of the original position istroubling. Only if Individuals are rational do their negotiations andprinciples merit our attention, let alone our acceptance. To avoid intro-ducing into the concept of rationality any controversial ethical state-ments which would jeopardize wide acceptance of the original positionas a satisfactory analytic construct, rationality is interpreted as takingthe most effective means to given ends. Neither the interpretation nor itsrationale are sound.

Throughout A Theory of Justice Rawls assumes that if everyonewould agree to an interpretation or principle when veiled in ignorance,

7 Teleological theories assert that the only consideration which mahes anaction or rule right or obligatory is the goodness of its consequences; deontologicaltheories claim that certain features of the act itself other than the value it bringsabout must be considered. For a dear, brief introduction to ethics and kinds ofethical theories, see W. Frankena, Ethics (1963).

Imaged with the Permission of N.Y.U. Law Review

November 19721

NEW YORK UNIVERSITY LAW REVIEW [Vol. 47:1012

then once that veil lifts, one can only supplement, but never radicallyalter or replace, that interpretation or principle. Principles of riskaccepted in the original position, for instance, are carried over intactto situations where one does not labor under full ignorance. Yet surelywhat might be agreed to under Rawls's special, stipulated conditionswould likely be uncompelling or even irrational under normal condi-tions; policies which are reasonable under conditions of limited knowl-edge may well be unreasonable given substantially more knowledge. Byholding Individuals to what they agree to when veiled in ignorance,Rawls no doubt avoids messy disputes regarding the relevance andweight of various considerations-but only at the cost of distorting theconcept of rationality.

Rationality interpreted as efficiency may sound plausible when endsare not in dispute. Unfortunately, this interpretation glosses over thefact that people have a multiplicity of ends ordered and related in com-plex and incompatible ways. Nor can a sharp distinction be drawnbetween means and ends; means and ends are so related that, exceptin the simplest cases, an alteration of one affects the nature of the other.One cannot suppose, for example, that every effective means of teachingpeople always results in their having the same education; a student whostudies Shakespeare by means of a programmed text is likely to have adifferent understanding and appreciation of the subject than one wholearns Shakespeare through the theater, though both students maypossess the same facts.

Rawls's interpretation of rationality is defective in another way.Many people value certain ends above any purely formal procedureof decisionmaking, yet Rawls would have them sacrifice these ends for aprocedure they would have adopted had they been Rawls's RationalIndividuals. If there is an end that a man is convinced (rightly orwrongly) requires the complete devotion of everyone, he would beunlikely to limit his pursuit of this end whenever it ran afoulwith a procedure he did not in fact adopt. There seems to be no proce-dural shortcut which will eliminate conflicts over ends or, indeed, entireways of life. Raws, we will recall, contends that purely proceduraljustice is rational "when there is no independent criterion for the rightresult" (p. 86). Perhaps, but this will not persuade those who believethat such a criterion exists and, what is more, that they know what it is.That there is great disagreement concerning the nature of this criterionis neither here nor there; a person who is convinced that a certainarrangement is just will not usually be moved to adopt a principle ofpure procedural justice merely because others disagree with him. Ra-tional Individuals adopt a purely procedural principle because theyhave no deep moral and value convictions; real people who have suchconvictions will resist suppositions which require that they set theseconvictions aside. Rawls forbids his Rational Individuals to have a fulltheory of the good because he does not wish to jeopardize the prior

Imaged with the Permission of N.Y.U. Law Review

BOOK REVIEWS

place of the concept of right, but this unfairly tips the scale againstteleological views at the outset.

The very intelligibility of the original position is not beyondquestion. Can Individuals be unaware of the civilization to which theybelong, yet cognizant of the laws of economics, the principles of psycho-logical development and the nature of political affairs? Because physicallaws are unaffected by cultural factors, may we assume that the same istrue of psychological, sociological and economic laws? Once Rawlsspells out the nature of these laws in detail, as he must, one may findthat they are only applicable to some societies at a certain stage of theirhistorical development. Further, if man is necessarily a social being, i.e,a creature whose nature cannot be adequately understood apart from hissocial relationships, then to develop a theory of justice from the view-point of theoretically defined individuals living in splendid isolation fromany particular social and historical context is misguided. If JeremyBentham took human nature to be that of the English shopkeeper, asKarl Marx alleged, at least he took the point of view of actual peoplesituated in an actual historical situation; Rawls's Rational Individuals,on the other hand, correspond to no one in any social or historical sit-uation.

Behind these specific criticisms lies a nagging feeling that theoriginal position merely reflects, in an abstract way, contemporaryliberal democratic attitudes--the description of it has appeal preciselybecause it sums up our liberal, individualistic outlook. Perhaps this isinevitable; given Rawls's pragmatic approach we are bound to find aprominent place for our firmly held moral beliefs. Yet unless our atti-tudes and beliefs are grounded in more than consensus, their appealwill be less widespread than Rawls anticipates. My concern, to put thepoint differently, is that any principles derived from the original posi-tion will simply reproduce in theory what we heirs to the liberal, demo-cratic Western tradition intuitively believe in fact. Unless we are pro-vided with the metaphysics of a theory of justice, which Rawls studi-ously avoids, I do not see how Rawls can overcome this essentiallyHegelian objection.

Professor Rawls's decision to cast his theory in terms of an idealagreement or social contract is also troublesome. Specifically, Rawlsnever adequately describes the tasks the social contract model performsor how literally the model should be taken. Rawls dearly believes thatthe concept of social justice can be illuminatingly analyzed in terms ofa contract which Rational Individuals in the original position wouldunanimously make. Precisely what is shown, however, is never madedear. In particular, it is not dear whether the original position and thesocial contract are intended as devices (1) for analyzing a conception ofjustice we already have, (2) for generating principles of justice whichthen must be justified in noncontractarian ways or (3) for justifyingas well as generating principles of social justice.

Imaged with the Permission of N.Y.U. Law Review

November 1972]

NEW YORK UNIVERSITY LAW REVIEW [Vol. 47:1012

I am inclined to think that Rawls uses the original position andthe social contract simply as tools of analysis and devices for generatingprinciples of justice, and not as a way of justifying these principles. Ido so not because the weight of textual evidence clearly points in thisdirection, but because it seems obvious that hypothetical contractsagreed to by hypothetical individuals in a hypothetical situation cannotbind me. How could they, when what I myself would have done (butdid not) cannot bind me. Had a particular student asked me to directhis research, I might well have promised to do so; given that I made nosuch promise, however, I am now under no obligation to him. There.fore, if I am not bound by a hypothetical contract that I myself wouldhave made, I am surely not bound by a hypothetical contract thatcertain mythical creatures would make.

On this interpretation of Rawls, the social contract functions asa way of perceiving, conceiving and construing that part of moralityconcerned with justice. Although other perspectives abound, none is asfruitful. The concepts of equality and the fortuitousness of human differ-ences in nature and situation are among the central notions Rawls'saccount attempts to explain; these notions are also among our deepestmoral intuitions. The social contract itself may be an artificial notion.But, as David AJ. Richards, a proponent of Rawls's method, declares,

there is nothing artificial or superficial about the ideas which the idealcontract view is trying to articulate: these ideas are as real as our oldestand most haunting collective myths. The point of the theoretical elabo-rateness of the ideal contract view is not artificially to sophisticate theseprofound and durable ideas, but rather to provide or suggest some wayin which these ideas may be so expressed that one may fruitfully indi-cate how these notions establish the specific and determinate structureof our moral experience. The notion of contractors deciding what countas moral principles must be understood from this point of view: as away of giving workable expression to certain deep and permanent featuresof moral thought .... 8

Like any other perspective, that of the social contract must be judgedby its fruits. To Richards this means that the social contract model mustultimately be judged

in terms of the creative usefulness of such theoretical intricacy in pro-viding tools with which we may take hold and better understand not onlythe nature of some of the most abstract and difficult concepts which weemploy, but also bow such concepts are specifically applied to the varie-gated circumstances of human life.9

I have spoken of the social contract as a model from which one canrigorously deduce principles of justice, and certainly there are passagesto justify this interpretation. Rawls claims, for example, that his argu-ment "aims eventually to be strictly deductive . .. We should strive

8 D. Richards, A Theory of Reasons for Action 91 (1971). Rawls supervisedRichards's undergraduate senior thesis, and Richards acknowledges the considerableinfluence Rawls's own work has had on him; clearly, the influence is enormous.

9 Id.

Imaged with the Permission of N.Y.U. Law Review

BOOK REVIEWS

for a kind of moral geometry with all the rigor which this name con-notes" (p. 121). Other passages suggest that the social contract and theoriginal position are merely vivid metaphors -which play no essentialrole in the argument. Thus, we find Rawls asserting:

To say that a certain conception of justice would be chosen in the originalposition is equivalent to saying that rational deliberation satisfying cer-tain conditions and restrictions would reach a certain conclusion. Ifnecessary, the argument to this result could be set out more formally.I shall, however, speak throughout in terms of the notion of the originalposition. It is more economical and suggestive, and brings out certainessential features that otherwise one might easily overlook (p. 138 (em-phasis added); cf. pp. 21-22).

Whether model or metaphor, the social contract and the original positiongive rise to certain misgivings as regards their usefulness. Some of thesehave already been discussed; others will be mentioned as we examineRawls's principles of justice.

II1

Two PIcIPLES or SocIAL JusTicE

Rational Individuals in the original position, we are told, wouldadopt two principles of justice ranked in a certain order. According tothe principle of equal liberty, "[e]ach person is to have an equal rightto the most extensive total system of equal basic liberties compatiblewith a similar system of liberty for all" (p. 302). This principle, Rawlsargues, takes precedence over the principle regulating economic andsocial inequalities, for, assuming a minimum level of general well-being,it is "irrational from the standpoint of the original position to acknowl-edge a lesser liberty for the sake of greater material means and amenitiesof office" (p. 542). Economic and social benefits (and burdens) are tobe distributed in accordance with the principle of equal benefits andopportunities: "Social and economic inequalities are to be arranged sothat they are both: (a) to the greatest benefit of the least advantaged ...and (b) attached to offices and positions open to all under conditions offair equality of opportunity" (p. 302).

Assuming a minimum level of general well-being, liberty can berestricted only if a less extensive liberty strengthens the total Oharedsystem of liberty or if unequal liberty is acceptable to those who willhave less liberty. Liberty of conscience and the rights which define theintegrity of the person are our fundamental basic liberties; certainpolitical rights, on the other hand, may be foregone either to strengthenthe entire system of justice or to compensate those less fortunate (ci.p. 247). Yet even these inequalities are justified only as long as socialconditions are being developed under which less than equal libertywould be unjustified. Raws's purpose in ordering the two principles ashe does is to prohibit exchanges between basic liberties and economicor social benefits. If minimum human wants are satisfied, economicconsiderations cannot justify suspension of any basic liberties, including

Imaged with the Permission of N.Y.U. Law Review

November 1972]

NEW YORK UNIVERSITY LAW REVIEW [Vol, 47:1012

the right to vote and to hold public office, freedom of speech and as-sembly and the right to hold personal (not necessarily private) property.

Once the principle of equal liberty is satisfied, one might betempted to adopt the principle of Pareto optimality to determine a justdistribution of goods. According to this principle, "a configuration isefficient whenever it is impossible to change it so as to make somepersons (at least one) better off without at the same time making otherpersons (at least one) worse off" (p. 67). Although Rational Individualswould accept this principle to judge the efficiency of their economic andsocial arrangements, Rawls argues convincingly that they would rejectit as a principle of justice, since there are usually many arrangementswhich, though efficient, are manifestly unjust-including the limitingcase where one person (or class) receives every available benefit.

If Rational Individuals would reject the principle of efficiency, sowould they also reject any utilitarian principle. Classical utilitarianism,which bids us to maximize total (net) benefits, permits and even re-quires some people to carry enormous burdens provided only that theirsacrifices maximize utility over the whole society. Since we are assumingthat utility cannot serve as the first principle of justice and that theprinciple of equal liberty, which is the first principle, has been satisfied,the worst cases of injustice could not occur. Even so, there would seemto be no reason why a Rational Individual would choose a principle ofdistribution which might well be to his disadvantage however much itmight benefit society as a whole. For similar reasons, Rational Indi-viduals would reject suggestions that average (net) utility be maximized.

Rational Individuals in the original position, Rawls argues, wouldadopt the difference principle. Of the various efficient arrangements, thedifference principle singles out that distribution of benefits which im-proves the expectations of the least advantaged members of society:

Assuming the framework of institutions required by equal liberty andfair equality of opportunity, the higher expectations of those bettersituated are just if and only if they work as part of a scheme whichimproves the expectations of the least advantaged members of society.(p. 75).

Admittedly, not all arrangements satisfying this principle will maximizetotal or even average utility. However, Rational Individuals wouldchoose a principle promising less than the utility principle because, asrational creatures, they would adopt a maximin policy; that is, a policyof accepting only those principles which have the best worst result.People who wear raincoats on cloudy days act on this policy-though anuisance, a raincoat guarantees the best result (keeping dry) should theworst conditions obtain (rain), though at the cost of forfeiting one'schances for the best possible outcome (neither rain nor raincoat). AsRational Individuals have no idea of their eventual position in society,they would subscribe to a maximin rule. This rule, though not part ofRawls's definition of rationality, is presumed to guide the choice of bothprinciples of justice.

Imaged with the Permission of N.Y.U. Law Review

BOOK REVIEWS

By distinguishing between equal liberty and opportunity on theone hand and the worth of equal liberty and opportunity on the otherhand, Rawls acknowledges that poverty, ignorance and natural liabilitiesmay render formal equality virtually worthless. However, I am not surethat Rawls traces the implications of this distinction correctly. If every-one has equal liberty, the question of compensation for lesser libertyobviously does not arise; but how are we to compensate the disadvan-taged for the lesser worth of liberty? "The lesser worth of liberty is ...compensated for, since the capacity of the less fortunate members ofsociety to achieve their aims would be even less were they not to acceptthe existing inequalities whenever the difference principle is satisfied"(p. 204). This is true but irrelevant. The disadvantaged would indeedbe ill-advised to take less than the difference principle authorizes giventhat equality of liberty must be iunintained. But this begs the question.The disadvantaged might argue that, because their basic liberties areworth so little, they ought to be given greater liberties and rights thanthe advantaged enjoy, e.g., double votes. Just this sort of argument hasin fact been advanced by radical blacks in an effort to secure automaticrepresentation on important decisionmaking bodies. The argument maybe unsound, unjust or both, but Rawls does not show that it is. Onemight even argue that Rational Individuals obeying a maximin rulewould agree to inequalities of basic liberties in favor of disadvantagedgroups whenever the worth of liberty falls below a certain level or when-ever certain human needs are being ignored by existing social institu-tions; after all, such inequalities would seem to maximize one's well-being were one to end up among the least advantaged.

The maximin rule itself needs clarification and support. In explain-ing the rule, Rawls says that "the two principles are those a personwould choose for the design of a society in which his enemy is to assignhim his place" (p. 152). Were one's position indeed determined byone's enemies, Rawls's defense of the principles makes sense. However,Brian Barry points out that "our hypothetical prindple-choosers arenot going to be assigned their positions by their enemy."10 Rather, theirpositions will depend on personal characteristics and good fortune. Thisplay-safe strategy is therefore unnecessary.

Consider again the man who carries a raincoat on cloudy days.If he believes that there exists a malevolent being intent on causinghim to suffer, then of course the maximin rule is reasonable; if hebelieves on the other hand that there exists a benevolent being intenton causing him to prosper, then a maximax rule, prescribing that heopt for the best possible outcome, is reasonable. But supposing, ashe surely would, that the weather is not affected by his deliberations,our rational man would estimate the likelihood of rain and dressaccordingly. Individuals pursuing their own interests exclusively mightchoose those principles leading to a high average level of well-being;"whether they would prefer one with more equal or less equal dis-

10 Barry, On Social Justice, Oxford Rev. 36 (Trinity Term 1967).

Imaged with the Permission of N.Y.U. Law Review

November 1972]

NEW YORK UNIVERSITY LAW REVIEW [Vol. 47:1012

tribution," Barry concludes, "would depend on their taste for gam-bling.""

Rawls introduces the veil of ignorance just to meet such objec-tions; under conditions of ignorance, insists Rawls, Rational Individ-uals could not play the percentages and would therefore adopt themaximin rule. But would they? Wouldn't it be at least as rationalto agree to a social minimum and then devise basic institutions whichwould permit extreme inequalities? For no matter where one endedup in society, one would have a roof over one's head and other neces-sary creature comforts. Once this floor had been decided upon, how-ever, Rational Individuals might choose to gamble for stakes involvingsubstantial inequalities, especially if the inequalities were thought tobe necessary to purchase a high level of art, learning or even a hedonicparadise.

However, supposing that Rational Individuals would subscribeto a maximin rule, Rawls stops short of its full implications. He as-sumes without argument that Rational Individuals would choose prin-ciples applicable only to representative persons, not to particular oractual persons. Yet if Rational Individuals are completely self-inter-ested, Rawls cannot suppose that they would be concerned with theinterests of any representative or typical or average person of anygroup; for no one knows whether he will be representative, typicalor average. Thus the expectations of representative persons are ir-relevant. To use Rawls's own heuristic device, imagine that one'senemy were to assign one's place in society. Adopting the maximinrule, everyone would choose only those principles guaranteeing thatany economic or social distribution must improve the expectationsof each and every member of the least-advantaged group. Rawls'sattempt to avoid the strict egalitarian implications of the maximinprinciple therefore fails.

These criticisms make apparent that the social contract concep-tion of justice can generate any number of principles depending onthe conditions built into the interpretation of the original negotiatingposition. Why should we accept the conditions Rawls imposes onthe initial choice situation? Rawls addresses this question twice, onceat the beginning of A Theory of Justice and once at the very end;both times the answer is the same. Despite the hypothetical natureof the original position and of the social contract,

the conditions embodied in the description of the original position areones that we do in fact accept. Or if we do not, then perhaps wecan be persuaded to do so by philosophical reflection. Each aspect of thecontractual situation can be given supporting grounds.... (p. 21; seep. 587).

"Philosophical reflection" appears to refer to Rawls's belief thatmoral philosophy is Socratic:

[W]e may want to change our present considered judgments once their11 Id. at 38.

Imaged with the Permission of N.Y.U. Law Review

BOOK REVIEWS

regulative principles are brought to light. And we may want to do thiseven though these principles are a perfect fit. A knowledge of theseprinciples may suggest further reflections that lead us to revise ourjudgments.... (p. 49).

The resulting reflective equilibrium is therefore much more thanmatching one's existing judgments with moral principles; the kindof reflective equilibrium which concerns moral philosophy involves aninvestigation of various conceptions of justice and a consideration ofphilosophically relevant arguments on behalf of each. The originalposition and the social contract, therefore, are e.xpository deviceswhich draw attention to conditions we either accept or can be broughtto accept.

Rawls stresses that his theory of justice is a theory of the moralsentiments; that is, it sets out the principles governing our moralcapacities and powers, espedally our sense of justice. Limitationsof space prevent a detailed exposition and evaluation of thisaspect of Rawls's theory. I mention it, however, because of its bear-ing on the question of justification of the two principles of justice.We have already seen that definitions and analyses of meaning haveno special place in Rawls's approach; they stand or fall with thetheory itself. The same is true, or is suggested to be true, of justifica-tion. "[I] f we can find an accurate account of our moral conceptions,"Rawls declares, "then questions of meaning and justification mayprove much easier to answer. Indeed," he concludes, "some of themmay no longer be real questions at all" (p. 51). Just as developmentssince Frege and Cantor have immeasurably deepened our understand-ing of the meaning and justification of statements in logic and mathe-matics, so Rawls hopes that if he can provide an adequate theoryof our moral sentiments and concepts, a similar transformation mayoccur in ethics.

IV

CONCLUSION

Readers of this journal -will find that A Theory of Justice il-luminates a variety of topics and controversies I have slighted orignored altogether: the nature and justification of conscientious refusal,civil disobedience and political obligation; the rule of law and thestatus of majority rule; the limits of tolerance and conscience; guilt,shame and happiness; duties, natural duties and obligation; and thenature of a well-governed society. Rawls's analyses are sensible, subtleand occasionally yield surprising conclusions. He suggests, for instance,that citizens generally (as opposed to officeholders) have no politicalobligation (see p. 114) and that the good things of life should notbe distributed according to moral desert (see p. 310).

There is much for everyone to learn from this book, though eachreader will have his own reservations and concerns. But this much isbeyond doubt: A Theory of Justice makes an outstanding contribu-

Imaged with the Permission of N.Y.U. Law Review

November 19721

NEW YORK UNIVERSITY LAW REVIEW

tion to moral, social and political thought; it deserves its inevitablefate of being the most discussed book in its field for the next decade.

To see our place in society from the perspective of contractorsin the original position, Raws concludes, is to see it sub spccieaeternitatis: "it is to regard the human situation not only from allsocial but also from all temporal points of view" (p. 587). If ATheory of Justice actually provides us with a temporal perspectivetied to liberal, democratic preconceptions, a perspective which mayalready be on the wane, we may well recall Hegel's observation thatthe owl of Minerva spreads its wings only at dusk.

HANS OBERDIE0'

* Professor of Philosophy, Swarthmore College.

Imaged with the Permission of N.Y.U. Law Review